While studies have used acoustic sensors to determine fracture initiation time in biomechanical studies, a systematic procedure is not established to process acoustic signals. The objective of the study was to develop a methodology to condition distorted acoustic emission data using signal processing techniques to identify fracture initiation time. The methodology was developed from testing a human cadaver lumbar spine column. Acoustic sensors were glued to all vertebrae, high-rate impact loading was applied, load-time histories were recorded (load cell), and fracture was documented using CT. Compression fracture occurred to L1 while other vertebrae were intact. FFT of raw voltage–time traces were used to determine an optimum frequency range associated with high decibel levels. Signals were bandpass filtered in this range. Bursting pattern was found in the fractured vertebra while signals from other vertebrae were silent. Bursting time was associated with time of fracture initiation. Force at fracture was determined using this time and force–time data. The methodology is independent of selecting parameters a priori such as fixing a voltage level(s), bandpass frequency and/or using force–time signal, and allows determination of force based on time identified during signal processing. The methodology can be used for different body regions in cadaver experiments.